The present invention relates to therapeutic polypeptides useful, e.g., for the treatment of neurological and other disorders.
Insulin-like growth factors (IGFs) have been identified in various animal species as polypeptides that act to stimulate growth of cells in a variety of tissues (see Baxter et al., Comp. Biochem. Physiol. 91B:229-235 (1988); and Daughaday et al., Endocrine Rev. 10:68-91 (1989) for reviews), particularly during development (see D'Ercole, J. Devel. Physiol. 9:481-495 (1987) for review). The IGFs, each of which has a molecular weight of about 7,500 daltons, are chemically related to human proinsulin: i.e. they possess A and B domains that (1) are highly homologous to the corresponding domains of proinsulin, and (2) are connected by a smaller and unrelated C domain. A carboxyl-terminal extension, the D domain, is also present in IGFs but is not found in proinsulin.
Certain polypeptide fragments of the IGFs have proven to be useful as antigens to raise antibodies specific for each of the IGFs (see, e.g., Japanese Patent Application No. 59065058; Hintz and Liu, J. Clin. Endocr. Metab. 54:442-446 (1982); Hintz et al., Horm. Metab. Res. 20:344-347 (1988)). Using labelled IGF-specific antibodies as a probe, IGF-I and IGF-II (sometimes respectively termed "somatomedin C" and "somatomedin A") have been found in a variety of tissues, including the mammalian central nervous system(CNS); the presence in the CNS of mRNAs encoding these polypeptides suggests local synthesis in the CNS (see Baskin et al., TINS 11:107-111 (1988) for review). In addition, IGF-III (or "brain IGF"), a truncated form of IGF-I lacking the latter protein's three N-terminal amino acid residues, has been found in fetal and adult human brain (Sara et al., Proc. Natl. Acad. Sci. USA 83:4904-4907 (1986), as well as in colostrum (Francis et al., Biochem. J. 251:95-103 (1988)). Two different IGF receptors have been identified in the adult human CNS (Baskin et al., 1988, supra), including in the brain (Sara et al., Neurosci. Let. 34:39-44 (1982)). In addition, European Patent No. 227,619 describes evidence for a third type of IGF receptor located in human fetal membranes. Complicating research in this area are (1) evidence that the insulin receptor of brain membranes recognizes not only insulin but also the IGFs; (2) the finding that one of the two types of adult IGF receptors exhibits some affinity for insulin as well as for both IGF-I and II, and (3) current uncertainty as to the physiological significance of binding of IGF-II to the second type of adult IGF receptor (Baskin et al., 1988, supra).
IGF-I and IGF-II appear to exert a stimulatory effect on development or proliferation of a wide range of susceptible cell types (see Daughaday et al., 1989, supra, for review). Treatment with the IGFs or with certain polypeptide fragments thereof has been variously suggested as a bone repair and replacement therapy (European Patent Application No. 289 314), as a means to counteract certain harmful side effects of carcinostatic drugs (Japanese Patent No. 63196524), and as a way to increase lactation and meat production in cattle and other farm animals (Larsen et al., U.S. Pat. No. 4,783,524). Each of the IGFs also appears to enhance the survival, proliferation and/or neurite outgrowth of cultured embryonic neurons (which, unlike mature neurons, have not yet lost their ability to undergo cell division) from various parts of the CNS (Aizenman et al., Brain Res. 406:32-42 (1987); Fellows et al., Soc. Neurosci. Abstr. 13:1615 (1987); Onifer et al., Soc. Neurosci. Abstr. 13:1615 (1987); European Patent No. 227,619 and from the peripheral nervous system (Bothwell, J. Neurosci Res. 8:225-231 (1982); Recio-Pinto et al., J. Neurosci 6:1211-1219 (1986)). In addition, the IGFs have been shown to affect the development of undifferentiated neural cells: human neuroblastoma tumor cells were shown to respond to added IGFs by extending neurites (Recio-Pinto and Ishii, J. Neurosci. Res. 19:312-320 (1988)) as well as by undergoing mitosis (Mattson et al., J. Cell Biol. 102:1949-54 (1986). As the induction of the enzyme ornithine decarboxylase has been shown to correlate with the stimulation of mitotic activity of these cells, an assay for cell proliferation has been developed based upon measuring the level of activity of this enzyme (Mattsson et al., 1986).
Developing forebrain cholinergic neurons (cultured rat septal neurons) are sensitive to a variety of growth factors in vitro. Addition of nerve growth factor (NGF) to the culture medium increases the number of cells positive for the expression of transmitter-specific enzymes (acetyl choline esterase (ACHE) and choline acetyl transferase (CHAT)) (Hartikka and Hefti, J. Neuroscience 8:2967-2985 (1988). Thyroid hormone also increases the level of ChAT in cultured septal neurons and thyroid hormone in combination with NGF results in a stimulation of ChAT activity much greater than the sum of the effects of individual addition of these two substances (Hayashi and Patel, Dev. Brain Res. 36:109-120 (1987)). IGF-I, IGF-II, and insulin also induce ChAT activity in cultured septal neurons (Knusel et al., J. of Neuroscience 10:558-570 (1990)). When NGF and insulin are both added to the culture medium the effect on ChAT activity is additive, but the effects of IGF-I or IGF-II in combination with insulin are not additive (Knusel et al., 1990, supra).
In vivo studies also support the hypothesis that the IGFs play a role in development and differentiation of the immature peripheral and central nervous systems (Sara et al., J. Dev. Physiol. 1:343-350 (1979); Philipps et al., Pediatr. Res. 23:298-305 (1988); Sara et al., Prog. Brain Res. 73:87-99 (1988)), although the physiological nature of this role remains uncertain. Once the neuronal cells of the CNS reach maturity, they do not undergo further cell division.
Neurotrophic factors other than the IGFs have been proposed as a potential means of enhancing neuronal survival, for example as a treatment for the neurodegenerative diseases amyotrophic lateral sclerosis (using skeletal muscle-derived proteins having apparent molecular weights in the 20,000-22,000 dalton and 16,000-18,000 dalton ranges: PCT Application No. PCT/US88/01393), and Alzheimer's disease (using phosphoethanolamine: PCT Application No. PCT/US88/01693). Sara et al., although finding a "significant elevation" in serum and cerebrospinal fluid somatomedin (IGF) levels in patients suffering from Alzheimer's disease compared to normal controls, nevertheless conclude:
Whether somatomedins play a casual (sic) role in the etiology of the dementia disorders of the Alzheimer type remains to be determined. However, since somatomedins stimulate the uptake of amino acids into brain tissue, their administration may provide beneficial therapeutic effects. Finally, the fall in somatomedins observed in normal elderly patients raises the general question of their role in cell aging. (citation omitted; Sara et al., Neurobiol. Aging 3:117-120, 119 (1982)). PA1 [One] of the major deficits in Alzheimer's disease concerns the cholinergic system of the brain, where a reduced synthesis and release of [acetylcholine] has been found . . . . It is of considerable importance to further investigate the role of IGFs in neurodegenerative disorders such as Alzheimer's disease . . . (citations omitted). PA1 Thus, increased IGF-I immunoreactivity is observed in regenerating peripheral nerves after any injury and seems to form part of a general reaction pattern, most evident in the Schwann cells. Our ultrastructural studies have revealed that the Schwann cells undergo hypertrophy after vibration trauma, and show signs of activation, i.e. the granular endoplasmic reticulum and Golgi complex increased in extent. We thus interpret the increase in IGF-I immunoreactivity in the Schwann cells, documented in this study on vibration-exposed nerves, as part of a transient, reactive response beneficial for the early stages of repair processes . . . . We consider the increase in IGF-I immunoreactivity to reflect mainly the initial reactions in a chain of events resulting in repair of the injured tissue or organ [although this increase] may be interpreted to reflect disturbed axoplasmic transport [of IGF-I molecules], due in part to the diminution of microtubules reported to occur after vibration exposure. (citation omitted)
In a report that IGF-I, but not IGF-II, stimulates the immediate (i.e. within 20 min.) release of acetylcholine from slices of adult rat brain, a process thought to be related to transitorily increased neurotransmission of acetylcholine rather than to increased cholinergic enzyme activity, Nilsson et al., Neurosci. Let. 88:221-226, 221, 224 (1988), point out that
Using antibody specific for IGF-I to detect an increase in the presence of IGF-I in injured peripheral nerves, notably in the non-neuronal cells named "Schwann cells", Hansson et al., Acta Physiol. Scand. 132:35-41, 38, 40 (1988), suggest that
Further, Sjoberg et al., Brain Res. 485:102-108 (1989), have found that local administration of IGF-I to an injured peripheral nerve stimulates regeneration of the nerve as well as proliferation of associated non-neuronal cells.
Several methods have been employed to decrease the susceptibility of polypeptides to degradation by peptidases, including, e.g., substitution of D-isomers for the naturally-occurring L-amino acid residues in the polypeptide (Coy et al., Biochem. Biophys. Res. Commun. 73:632-8 (1976)). Where the polypeptide is intended for use as a therapeutic for disorders of the CNS, an additional problem must be addressed: overcoming the so-called "blood-brain barrier," the brain capillary wall structure that effectively screens out all but selected categories of molecules present in the blood, preventing their passage into the brain. While the blood-brain barrier may be effectively bypassed by direct infusion of the polypeptide into the brain, the search for a more practical method has focused on enhancing transport of the polypeptide of interest across the blood-brain barrier, such as by making the polypeptide more lipophilic, by conjugating the polypeptide of interest to a molecule which is naturally transported across the barrier, or by reducing the overall length of the polypeptide chain (Pardridge, Endocrine Reviews 7:314-330 (1986); U.S. Pat. No. 4,801,575.